# Arrays

# Create Array of Strings

Arrays of strings can be created using ruby's percent string (opens new window) syntax:

array = %w(one two three four)

This is functionally equivalent to defining the array as:

array = ['one', 'two', 'three', 'four']

Instead of %w() you may use other matching pairs of delimiters: %w{...}, %w[...] or %w<...>.

It is also possible to use arbitrary non-alphanumeric delimiters, such as: %w!...!, %w#...# or %w@...@.

%W can be used instead of %w to incorporate string interpolation. Consider the following:

var = 'hello'

%w(#{var}) # => ["\#{var}"]
%W(#{var}) # => ["hello"]

Multiple words can be interpreted by escaping the space with a .

%w(Colorado California New\ York) # => ["Colorado", "California", "New York"]

# Create Array with Array::new

An empty Array ([]) can be created with Array's class method, Array::new:

Array.new

To set the length of the array, pass a numerical argument:

Array.new 3 #=> [nil, nil, nil]

There are two ways to populate an array with default values:

Pass an immutable value as second argument.
Pass a block that gets current index and generates mutable values.

Array.new 3, :x #=> [:x, :x, :x]

Array.new(3) { |i| i.to_s } #=> ["0", "1", "2"]

a = Array.new 3, "X"            # Not recommended.
a[1].replace "C"                # a => ["C", "C", "C"]

b = Array.new(3) { "X" }        # The recommended way.
b[1].replace "C"                # b => ["X", "C", "X"]

# Create Array of Symbols

array = %i(one two three four)

Creates the array [:one, :two, :three, :four].

Instead of %i(...), you may use %i{...} or %i[...] or %i!...!

Additionally, if you want to use interpolation, you can do this with %I.

a = 'hello'
b = 'goodbye'
array_one = %I(#{a} #{b} world)
array_two = %i(#{a} #{b} world)

Creates the arrays: array_one = [:hello, :goodbye, :world] and array_two = [:"\#{a}", :"\#{b}", :world]

# Manipulating Array Elements

Adding elements:

[1, 2, 3] << 4
# => [1, 2, 3, 4]

[1, 2, 3].push(4)
# => [1, 2, 3, 4]

[1, 2, 3].unshift(4)
# => [4, 1, 2, 3]

[1, 2, 3] << [4, 5]
# => [1, 2, 3, [4, 5]]

Removing elements:

array = [1, 2, 3, 4]
array.pop
# => 4
array
# => [1, 2, 3]

array = [1, 2, 3, 4]
array.shift
# => 1
array
# => [2, 3, 4]

array = [1, 2, 3, 4]
array.delete(1)
# => 1
array
# => [2, 3, 4]

array = [1,2,3,4,5,6]
array.delete_at(2) // delete from index 2
# => 3  
array 
# => [1,2,4,5,6] 


array = [1, 2, 2, 2, 3]
array - [2]
# => [1, 3]    # removed all the 2s
array - [2, 3, 4]
# => [1]       # the 4 did nothing

Combining arrays:

[1, 2, 3] + [4, 5, 6]
# => [1, 2, 3, 4, 5, 6]

[1, 2, 3].concat([4, 5, 6])
# => [1, 2, 3, 4, 5, 6]

[1, 2, 3, 4, 5, 6] - [2, 3]
# => [1, 4, 5, 6]

[1, 2, 3] | [2, 3, 4]
# => [1, 2, 3, 4]

[1, 2, 3] & [3, 4]
# => [3]

You can also multiply arrays, e.g.

[1, 2, 3] * 2
# => [1, 2, 3, 1, 2, 3]

# Creating an Array with the literal constructor [ ]

Arrays can be created by enclosing a list of elements in square brackets ([ and ]). Array elements in this notation are separated with commas:

array = [1, 2, 3, 4]

Arrays can contain any kind of objects in any combination with no restrictions on type:

array = [1, 'b', nil, [3, 4]]

# Accessing elements

You can access the elements of an array by their indices. Array index numbering starts at 0.

%w(a b c)[0] # => 'a'
%w(a b c)[1] # => 'b'

You can crop an array using range

%w(a b c d)[1..2] # => ['b', 'c'] (indices from 1 to 2, including the 2)
%w(a b c d)[1...2] # => ['b'] (indices from 1 to 2, excluding the 2)

This returns a new array, but doesn't affect the original. Ruby also supports the use of negative indices.

%w(a b c)[-1] # => 'c'
%w(a b c)[-2] # => 'b'

You can combine negative and positive indices as well

%w(a b c d e)[1...-1] # => ['b', 'c', 'd']

Other useful methods

Use first to access the first element in an array:

[1, 2, 3, 4].first # => 1

Or first(n) to access the first n elements returned in an array:

[1, 2, 3, 4].first(2) # => [1, 2]

Similarly for last and last(n):

[1, 2, 3, 4].last    # => 4
[1, 2, 3, 4].last(2) # => [3, 4]

Use sample to access a random element in a array:

[1, 2, 3, 4].sample  # => 3
[1, 2, 3, 4].sample  # => 1

Or sample(n):

[1, 2, 3, 4].sample(2) # => [2, 1]
[1, 2, 3, 4].sample(2) # => [3, 4]

# Arrays union, intersection and difference

x = [5, 5, 1, 3]
y = [5, 2, 4, 3]

Union (|) contains elements from both arrays, with duplicates removed:

x | y
=> [5, 1, 3, 2, 4]

Intersection (&) contains elements which are present both in first and second array:

x & y
=> [5, 3]

Difference (-) contains elements which are present in first array and not present in second array:

x - y
=> [1]

# Decomposition

Any array can be quickly decomposed by assigning its elements into multiple variables. A simple example:

arr = [1, 2, 3]
# ---
a = arr[0]
b = arr[1]
c = arr[2]
# --- or, the same
a, b, c = arr

Preceding a variable with the splat operator (*) puts into it an array of all the elements that haven't been captured by other variables. If none are left, empty array is assigned. Only one splat can be used in a single assignment:

a, *b = arr       # a = 1; b = [2, 3]
a, *b, c = arr    # a = 1; b = [2]; c = 3
a, b, c, *d = arr # a = 1; b = 2; c = 3; d = []
a, *b, *c = arr   # SyntaxError: unexpected *

Decomposition is safe and never raises errors. nils are assigned where there's not enough elements, matching the behavior of [] operator when accessing an index out of bounds:

arr[9000] # => nil
a, b, c, d = arr # a = 1; b = 2; c = 3; d = nil

Decomposition tries to call to_ary implicitly on the object being assigned. By implementing this method in your type you get the ability to decompose it:

class Foo
  def to_ary
    [1, 2]
  end
end
a, b = Foo.new # a = 1; b = 2

If the object being decomposed doesn't respond_to? to_ary, it's treated as a single-element array:

1.respond_to?(:to_ary) # => false
a, b = 1 # a = 1; b = nil

Decomposition can also be nested by using a ()-delimited decomposition expression in place of what otherwise would be a single element:

arr = [1, [2, 3, 4], 5, 6]
a, (b, *c), *d = arr # a = 1; b = 2; c = [3, 4]; d = [5, 6]
#   ^^^^^

This is effectively the opposite of splat.

Actually, any decomposition expression can be delimited by (). But for the first level decomposition is optional.

a, b = [1, 2]
(a, b) = [1, 2] # the same thing

Edge case: a single identifier cannot be used as a destructuring pattern, be it outer or a nested one:

(a) = [1] # SyntaxError
a, (b) = [1, [2]] # SyntaxError

When assigning an array literal to a destructuring expression, outer [] can be omitted:

a, b = [1, 2]
a, b =  1, 2  # exactly the same

This is known as parallel assignment, but it uses the same decomposition under the hood. This is particularly handy for exchanging variables' values without employing additional temporary variables:

t = a; a = b; b = t # an obvious way
a, b = b, a         # an idiomatic way
(a, b) = [b, a]     # ...and how it works

Values are captured when building the right-hand side of the assignment, so using the same variables as source and destination is relatively safe.

# #map

#map, provided by Enumerable, creates an array by invoking a block on each element and collecting the results:

[1, 2, 3].map { |i| i * 3 }
# => [3, 6, 9]

['1', '2', '3', '4', '5'].map { |i| i.to_i }
# => [1, 2, 3, 4, 5]

The original array is not modified; a new array is returned containing the transformed values in the same order as the source values. map! can be used if you want to modify the original array.

In map method you can call method or use proc to all elements in array.

# call to_i method on all elements
%w(1 2 3 4 5 6 7 8 9 10).map(&:to_i)
# => [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

# using proc (lambda) on all elements
%w(1 2 3 4 5 6 7 8 9 10).map(&->(i){ i.to_i * 2})
# => [2, 4, 6, 8, 10, 12, 14, 16, 18, 20]

map is synonymous with collect.

# Filtering arrays

Often we want to operate only on elements of an array that fulfill a specific condition:

# Select (opens new window)

Will return elements that match a specific condition

array = [1, 2, 3, 4, 5, 6]
array.select { |number| number > 3 } # => [4, 5, 6]

# Reject (opens new window)

Will return elements that do not match a specific condition

array = [1, 2, 3, 4, 5, 6]
array.reject { |number| number > 3 } # => [1, 2, 3]

Both #select and #reject return an array, so they can be chained:

array = [1, 2, 3, 4, 5, 6]
array.select { |number| number > 3 }.reject { |number| number < 5 }
  # => [5, 6]

# Inject, reduce

Inject and reduce are different names for the same thing. In other languages these functions are often called folds (like foldl or foldr). These methods are available on every Enumerable object.

Inject takes a two argument function and applies that to all of the pairs of elements in the Array.

For the array [1, 2, 3] we can add all of these together with the starting value of zero by specifying a starting value and block like so:

[1,2,3].reduce(0) {|a,b| a + b} # => 6

Here we pass the function a starting value and a block that says to add all of the values together. The block is first run with 0 as a and 1 as b it then takes the result of that as the next a so we are then adding 1 to the second value 2. Then we take the result of that (3) and add that on to the final element in the list (also 3) giving us our result (6).

If we omit the first argument, it will set a to being the first element in the list, so the example above is the same as:

[1,2,3].reduce {|a,b| a + b} # => 6

In addition, instead of passing a block with a function, we can pass a named function as a symbol, either with a starting value, or without. With this, the above example could be written as:

[1,2,3].reduce(0, :+) # => 6

or omitting the starting value:

[1,2,3].reduce(:+) # => 6

# Two-dimensional array

Using the Array::new (opens new window) constructor, your can initialize an array with a given size and a new array in each of its slots. The inner arrays can also be given a size and and initial value.

For instance, to create a 3x4 array of zeros:

array = Array.new(3) { Array.new(4) { 0 } }

The array generated above looks like this when printed with p:

[[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]

You can read or write to elements like this:

x = array[0][1]
array[2][3] = 2

# Arrays and the splat (*) operator

The * operator can be used to unpack variables and arrays so that they can be passed as individual arguments to a method.

This can be used to wrap a single object in an Array if it is not already:

def wrap_in_array(value)
  [*value]
end

wrap_in_array(1)
#> [1]

wrap_in_array([1, 2, 3])
#> [1, 2, 3]

wrap_in_array(nil)
#> []

In the above example, the wrap_in_array method accepts one argument, value.

If value is an Array, its elements are unpacked and a new array is created containing those element.

If value is a single object, a new array is created containing that single object.

If value is nil, an empty array is returned.

The splat operator is particularly handy when used as an argument in methods in some cases. For example, it allows nil, single values and arrays to be handled in a consistent manner:

def list(*values)
  values.each do |value|
    # do something with value
    puts value
  end
end

list(100)
#> 100

list([100, 200])
#> 100
#> 200

list(nil)
# nothing is outputted

# Turn multi-dimensional array into a one-dimensional (flattened) array

[1, 2, [[3, 4], [5]], 6].flatten  # => [1, 2, 3, 4, 5, 6]

If you have a multi-dimensional array and you need to make it a simple (i.e. one-dimensional) array, you can use the #flatten method.

# Get unique array elements

In case you need to read an array elements avoiding repetitions you case use the #uniq method:

a = [1, 1, 2, 3, 4, 4, 5]
a.uniq
#=> [1, 2, 3, 4, 5]

Instead, if you want to remove all duplicated elements from an array, you may use #uniq! method:

a = [1, 1, 2, 3, 4, 4, 5]
a.uniq!
#=> [1, 2, 3, 4, 5]

While the output is the same, #uniq! also stores the new array:

a = [1, 1, 2, 3, 4, 4, 5]
a.uniq
#=> [1, 2, 3, 4, 5]
a
#=> [1, 1, 2, 3, 4, 4, 5]

a = [1, 1, 2, 3, 4, 4, 5]
a.uniq!
#=> [1, 2, 3, 4, 5]
a
#=> [1, 2, 3, 4, 5]

# Get all combinations / permutations of an array

The permutation method, when called with a block yields a two dimensional array consisting of all ordered sequences of a collection of numbers.

If this method is called without a block, it will return an enumerator. To convert to an array, call the to_a method.

Example	Result
`[1,2,3].permutation`	`#<Enumerator: [1,2,3]:permutation`
`[1,2,3].permutation.to_a`	`[[1,2,3],[1,3,2],[2,1,3],[2,3,1],[3,1,2],[3,2,1]]`
`[1,2,3].permutation(2).to_a`	`[[1,2],[1,3],[2,1],[2,3],[3,1],[3,2]]`
`[1,2,3].permutation(4).to_a`	`[]` -> No permutations of length 4

The combination method on the other hand, when called with a block yields a two-dimensional array consisting of all sequences of a collection of numbers. Unlike permutation, order is disregarded in combinations. For example, [1,2,3] is the same as [3,2,1]

Example	Result
`[1,2,3].combination(1)`	`#<Enumerator: [1,2,3]:combination`
`[1,2,3].combination(1).to_a`	`[[1],[2],[3]]`
`[1,2,3].combination(3).to_a`	`[[1,2,3]]`
`[1,2,3].combination(4).to_a`	`[]` -> No combinations of length 4

Calling the combination method by itself will result in an enumerator. To get an array, call the to_a method.

The repeated_combination and repeated_permutation methods are similar, except the same element can be repeated multiple times.

For example the sequences [1,1], [1,3,3,1], [3,3,3] would not be valid in regular combinations and permutations.

Example	# Combos
`[1,2,3].combination(3).to_a.length`	1
`[1,2,3].repeated_combination(3).to_a.length`	6
`[1,2,3,4,5].combination(5).to_a.length`	1
`[1,2,3].repeated_combination(5).to_a.length`	126

# Remove all nil elements from an array with #compact

If an array happens to have one or more nil elements and these need to be removed, the Array#compact or Array#compact! methods can be used, as below.

array = [ 1, nil, 'hello', nil, '5', 33]

array.compact # => [ 1, 'hello', '5', 33]

#notice that the method returns a new copy of the array with nil removed,
#without affecting the original

array = [ 1, nil, 'hello', nil, '5', 33]

#If you need the original array modified, you can either reassign it

array = array.compact # => [ 1, 'hello', '5', 33]

array = [ 1, 'hello', '5', 33]

#Or you can use the much more elegant 'bang' version of the method

array = [ 1, nil, 'hello', nil, '5', 33]

array.compact # => [ 1, 'hello', '5', 33]

array = [ 1, 'hello', '5', 33]

Finally, notice that if #compact or #compact! are called on an array with no nil elements, these will return nil.

array = [ 'foo', 4, 'life']

array.compact # => nil

array.compact! # => nil

# Create an Array of consecutive numbers or letters

This can be easily accomplished by calling Enumerable#to_a (opens new window) on a Range object:

(1..10).to_a    #=> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

(a..b) means that it will include all numbers between a and b. To exclude the last number, use a...b

a_range = 1...5
a_range.to_a       #=> [1, 2, 3, 4]

('a'..'f').to_a    #=> ["a", "b", "c", "d", "e", "f"]
('a'...'f').to_a   #=> ["a", "b", "c", "d", "e"]

A convenient shortcut for creating an array is [*a..b]

[*1..10]           #=> [1,2,3,4,5,6,7,8,9,10]
[*'a'..'f']        #=> ["a", "b", "c", "d", "e", "f"]

# Create Array of numbers

The normal way to create an array of numbers:

numbers = [1, 2, 3, 4, 5]

Range objects can be used extensively to create an array of numbers:

numbers = Array(1..10) # => [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

numbers = (1..10).to_a # => [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

#step and #map methods allow us to impose conditions on the range of numbers:

odd_numbers = (1..10).step(2).to_a # => [1, 3, 5, 7, 9]

even_numbers = 2.step(10, 2).to_a # => [2, 4, 6, 8, 10]

squared_numbers = (1..10).map { |number| number * number } # => [1, 4, 9, 16, 25, 36, 49, 64, 81, 100]

All the above methods load the numbers eagerly. If you have to load them lazily:

number_generator = (1..100).lazy # => #<Enumerator::Lazy: 1..100>

number_generator.first(10) # => [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

# Cast to Array from any object

To get Array from any object, use Kernel#Array.

The following is an example:

Array('something') #=> ["something"]
Array([2, 1, 5])   #=> [2, 1, 5]
Array(1)           #=> [1]
Array(2..4)        #=> [2, 3, 4]
Array([])          #=> []
Array(nil)         #=> []

For example, you could replace join_as_string method from the following code

def join_as_string(arg)
  if arg.instance_of?(Array)
    arg.join(',')
  elsif arg.instance_of?(Range)
    arg.to_a.join(',')
  else
    arg.to_s
  end
end

join_as_string('something') #=> "something"
join_as_string([2, 1, 5])   #=> "2,1,5"
join_as_string(1)           #=> "1"
join_as_string(2..4)        #=> "2,3,4"
join_as_string([])          #=> ""
join_as_string(nil)         #=> ""

to the following code.

def join_as_string(arg)
  Array(arg).join(',')
end

# Syntax

a = [] # using array literal
a = Array.new # equivalent to using literal
a = Array.new(5) # create an array with 5 elements with value of nil.
a = Array.new(5, 0) # create an array with 5 elements with default value of 0.

← Comments Multidimensional Arrays →